JP6224068B2 - 3D display for multiple viewers - Google Patents

Description

  The present invention relates to an apparatus and method for processing three-dimensional (3D) image data. The present invention further emits at least two views of the 3D image data, the views through the left eye through the first view of the at least two views and through the right eye to provide a 3D effect. A 3D display that enables autostereoscopic viewing of the 3D image data at a plurality of viewing positions by a viewer who perceives a second of the at least two views, and display on the 3D display A display device comprising: a processor for processing the 3D image data to generate the at least two views for the purpose; and a viewer detector for detecting a viewer position of the viewer in front of the 3D display. .

  3D displays and in particular television receivers with 3D displays are becoming increasingly popular with consumers because they give viewers a stereoscopic perception of depth. Normally goggles are required, but viewers tend to prefer systems without additional glasses. Therefore, so-called autostereoscopic 3D displays have optical components, such as a lenticular lens array or a directional beam, to provide at least a left view and a right view of the scene. This allows the viewer to see different images with their own eyes if they are correctly positioned.

  Certain autostereoscopic displays, sometimes referred to as automultiscopic displays, provide multiple views of the same scene, rather than just left and right views. This allows the viewer to position in the conical field of view, i.e. move left and right in front of the display, while still acquiring stereoscopic perception of the scene.

  However, not all positions attached to the viewer are equally suitable for obtaining stereoscopic recognition of the scene. In particular, if the display is arranged to repeat the conical field of view as a series of conical fields, the viewer, for example, recognizes the outermost right view of a given conical field and the right eye May be positioned to recognize the outermost left view of an adjacent conical field of view. The viewer is then positioned at the turn of the cone. At such a viewing position, the viewer obtains a so-called pseudoscopic perception of the scene where the scene often appears to have depth but is inaccurate. Reverse visual acuity is known to cause headaches and other symptoms of visual fatigue.

  Regardless of the visual fatigue caused by reverse visual recognition, it is not always obvious to the viewer that the viewer is positioned at the reverse viewing position.

  US 6055013 discloses an autostereoscopic display that provides four views in each lobe, ie, in each conical field of view. An observer is said to have a wide range of viewing freedom with a reverse viewing position present at the boundary between adjacent lobes. The technique is represented to provide the viewer with a visual indication of the backsight position. To that end, the parallax optic is arranged to provide an indication in the dark central lobe, while the indication in the adjacent lobe is bright. The next lobe then darkens again and so on. Thus, at the reverse viewing position, at least one eye of the observer sees a position indication that is not black. As described above, the technique of Patent Document 1 enables an observer to distinguish between a desired stereoscopic view position and an undesired reverse view position.

U.S. Patent Application Publication No. 2004/0252187 describes a front projection display that can divide the viewer space into segments by using time sequential multiplexing. Each of the segments receives a different full resolution image stream. The system combines the pixel generation mechanism with an array of rotating micromirrors that sweep the pixels across user space to provide an operating mode for autostereoscopic 3D. The system comprises a viewer detector that determines the user's position relative to the screen, which position is obtained by head tracking. The location data is used to render a customized view for each particular user. Thus, the system generates multiple sequential view depths based on the user's position relative to the screen. In addition, a “fallback” scenario is described for the case where the user is not identified.
U.S. Patent Application Publication No. 2011/0316987 (Patent Document 3) displays a graphical symbol to indicate to a viewer whether they are located in a so-called orthoscopic area where a 3D effect exists. An autostereoscopic display system is described. Depending on the position of the viewer, the symbol can indicate to the viewer that he / she is not receiving the exact 3D effect.

US Pat. No. 6,055,033 US Patent Application Publication No. 2004/0252187 US Patent Application Publication No. 2011/0316987

  The problem with the above technique is that it is not suitable for providing a viewer with multiple views at any viewing position.

  An object of the present invention is a display device described in the section of the technical field, wherein at least a first viewer position of a first viewer is obtained via the viewer detector, and the first viewer The first viewer detects a viewer position conflict at the first viewer position where the first view and the second view do not bring the 3D effect to the first viewer. A viewer conflict detector arranged to control the generation of the view in dependence on the detected viewer position conflict by changing at least one of the at least two views received by A display device further comprising:

The viewer conflict detector uses the viewer detector to detect at least first and second viewers in front of the display. Thereafter, the viewer conflict detector determines at the viewer position whether the view provided to the first viewer provides an accurate 3D effect. Whenever it is determined that the viewer conflict detector does not provide the required 3D effect, the viewer conflict detector is connected to the processor and to the first viewer. Instructing to selectively change one of the views received by the first viewer to indicate that one viewer is not subject to the 3D effect, and the one selected view is the second Ensure that the view is not perceived by the viewer . For example, a warning can be dynamically inserted in one view that is recognized by the first viewer. Advantageously, the viewer conflict detector determines that one or more viewers are present at each position and further determines whether each of the viewers perceives the 3D effect. . Otherwise, the viewer conflict detector selectively changes one of the views to indicate, for example, to each viewer that the viewer will not have a 3D effect.

Optionally, the processor is arranged to perform the change by changing the one of the at least two views to indicate that the 3D effect is not brought about. This has the effect that other viewers who receive only the other one of the at least two views are not disturbed by the change.

  Optionally, the 3D display is a multi-view display arranged to emit adjacent to a series of more than two views of the 3D image data in a conical field of view, the series of views comprising a plurality of views in the conical field of view. Enabling autostereoscopic viewing of the 3D image data at a viewing position, wherein the viewer conflict detector is configured such that the first view is a first cone field from a first cone and the second view is It is arranged to detect a viewer position cone conflict at the first viewer position which is the second cone field from the second cone. Such a conical field of view is usually known to be adjacent to a further conical field of view by the mechanism that produces the cone described above, ie the lenticular lens array described above. On the boundary, the viewer can determine that the first view is a first cone field from a first cone and the second view is a second cone field to a second cone field. Recognize the reverse vision effect. Advantageously, the display device detects that the user is in such a position and changes the view.

Optionally, the viewer conflict detector detects the viewer position cone conflict when:
Overlaying graphical elements in each conical field of view,
Overlaying a text message on each conical field of view,
Reversing each conical field of view,
It is arranged to control the processor to change one of the first conical field and the second conical field by at least one of changing the image content of the respective conical field.

  Advantageously, the above changes provide a practical way to indicate to the viewer that the viewer is at a conflict location where the 3D effect is not exactly produced.

  Optionally, the viewer conflict detector detects the viewer position cone conflict to generate a 2D effect on the viewer to generate a 2D effect in the first cone field and the second cone field. Arranged to control the processor to change at least one of the first conical field of view and the second conical field of view by providing the same image content to both. Advantageously, viewers with viewer position cone conflicts no longer perceive the reverse vision effect and perceive the 2D effect. The lack of 3D effects can be just accepted by the viewer, or the viewer may move to another location.

Optionally, the viewer conflict detector detects the viewer position cone conflict when:
Changing the brightness of each conical field of view,
It is arranged to control the processor to change at least one of the first conical field and the second conical field by at least one of changing the color of each conical field.

  Advantageously, such changes are clearly visible to the viewer at the conflict location.

Optionally, the viewer conflict detector is
Obtaining at least a second viewer position of the second viewer via the viewer detector;
When detecting the viewer position cone conflict, one of the first cone field and the second cone field is selected, and the selected cone field is the second viewer. And is arranged to control the processor to change only the selected conical field of view.

  Advantageously, the second viewer is not affected by changes in the view.

Optionally, the 3D display emits a left view and a right view in succession via a respective first directional beam to the viewer's left eye and a respective second directional beam to the right eye. A sequential view display arranged such that the view enables autostereoscopic viewing of the 3D image data at each viewing position by combining the respective directional beams, and the viewer conflict detector
Obtaining at least a second viewer position of the second viewer via the viewer detector;
Arranged to detect a viewer position beam conflict where one view at the first viewer position coincides with one view at the second viewer position.

  The effect of sequential view, where each is directed to each eye of the viewer, is that the viewer recognizes the 3D effect without using any glasses and at the maximum resolution of the display. At this time, the viewer conflict detector determines the first and second viewer positions, and the left eye of the first user is in a position to receive the same view as the right eye of the second user. In some cases, the sequential view display cannot direct different views to their eyes. Thus, a viewer position beam conflict is detected when a particular view generated by one directional beam reaches two different respective eyes.

  Optionally, the viewer conflict detector detects the viewer position beam conflict to the other viewer of the first viewer and the second viewer. Provide the same image content for both the first view and the second view to provide a 2D effect to either the first viewer or the second viewer while providing a 3D effect Arranged to control the processor.

Optionally, the processor is
Preferences set by the first viewer or the second viewer for 3D or 2D mode;
The respective movements of the first viewer or the second viewer;
Based on at least one of the distances from the 3D display of each of the first viewer or the second viewer, either the first viewer or the second viewer One is arranged to provide the 2D effect.

  Advantageously, viewers who have a preferred or preferred position or may be aware of the 3D effect will recognize the 3D effect, while other users will recognize the 2D effect (eg, temporarily) To do.

Optionally, the processor is
Identification of each viewer and each viewer data Detection of facial features,
Eye distance detection,
The first view based on at least one of detection of the size of each of the face of the first viewer and the face of the second viewer detected via the viewer detector Based on the age of the viewer or the second viewer, it is arranged to bring the 2D effect to either the first viewer or the second viewer.

  The 3D effect is usually optimized for adult viewers. Children may have less optimal or distorted 3D effects, and some parents choose not to show 3D effects to their children. Advantageously, the system detects the viewer's age and selectively provides a 3D effect.

Optionally, the processor is arranged to generate at least three different sequential views, the sequential view display is arranged to emit the three sequential views in succession, and the first and second sequential views are both Constitutes a left view and right view pair, the second and third sequential views constitute a further left view and right view pair,
The viewer conflict detector controls the processor to provide the second sequential view as the matching view so that the first viewer and the second viewer are in the 3D. Arranged to recognize the effect. The sequential view display applies only two views, and a more flexible configuration is achieved by adding at least a third view so that the view pair is configured. Advantageously, the system may provide each viewer with a different pair of views to the viewer at the beam conflict location while sharing the second view with each left and right eye. Thus, every viewer experiences a 3D effect.

Optionally, the viewer detector is
Face detection in video images from cameras,
Distance information from the distance sensor,
User input from a remote control unit operated by the viewer at the viewer position,
Based on at least one of detecting a viewer tagged object positioned by the viewer at the viewer position, arranged to detect the viewer position of the viewer in front of the 3D display. The

  The viewer detector may determine the position of the viewer in the horizontal direction, i.e. in the angular direction and even in the distance, in order to determine the view that reaches each eye of the viewer. The viewer detector may include, for example, a distance sensor based on sound or infrared light. Such sensors may also provide both viewer distance and angle data. Also, the user may activate or position the remote control unit at his viewing position, or the user may define a predetermined viewing position that is predetermined. The viewer may also position the tagged object so that the viewer detector can accurately and easily detect the position of the viewer's eyes. For example, you may wear a visually detectable object around your neck.

In a further aspect of the invention, a method of processing 3D image data for a display device, wherein the display device emits at least two views of 3D image data, the view left so as to provide a 3D effect. At a plurality of viewing positions by a viewer perceiving a first view of the at least two views via the eye and a second view of the at least two views via the right eye. A 3D display that enables autostereoscopic viewing of the 3D image data; and a viewer detector that detects a viewer position of the viewer in front of the 3D display, the method comprising: Processing the 3D image data to generate the at least two views for display on a display, the method comprising:
Obtaining at least a first viewer position of the first viewer via the viewer detector;
Detecting a viewer position conflict at the first viewer position where the first view and the second view do not provide the 3D effect to the first viewer;
Obtaining at least a second viewer position of the second viewer via the viewer detector;
To indicate that the first viewer does not receive the 3D effect to the first viewer, selectively one of the at least two views received by the first viewer To generate the view depending on the detected viewer position conflict, so that the one selected view is not a view recognized by the second viewer. There is provided a method further comprising the step of detecting a viewer conflict by controlling.

  In a further aspect of the invention, a computer program is provided having instructions that cause a processor system to perform the above method. The computer program may be embodied on a computer readable medium.

  In accordance with the present invention, the processed 3D image data can be displayed on a 3D display that emits a series of views of the 3D image data, eg, an autostereoscopic multi-view 3D display. As such, the 3D image data may be so-called multi-view image data including image data for each view. Such 3D image data may be displayed directly on an autostereoscopic multiview display. Alternatively, the 3D image data may be stereoscopic image data including image data for the left view and the right view, or 2D image data plus depth data. Such 3D image data needs to be processed to obtain a series of views. Therefore, the 3D display may use so-called view interpolation in the case of stereoscopic image data, or so-called view rendering in the case of depth data added to 2D image data. It will be appreciated that the above techniques are known from the field of 3D image / video processing and in particular from the field of 3D image / video conversion.

  Such 3D displays are formed sequentially or in the form of a spatial conical field, i.e. using a directional beam or from the display surface of the 3D display, i.e. by the light emitting surface of the 3D display. A sequence of views is emitted as a sequence of views appearing conically from a surface. For example, the conical field of view of a 5-view autostereoscopic 3D display may have an outermost left view, an intermediate left view, a center view, an intermediate right view, and an outermost right view. When the viewer is located in a viewing position that recognizes two different views of the series of views, for example, the left eye is the outermost left view and the right eye is the middle left view, or the left When the middle view is recognized by the right eye and the middle right view is recognized by the right eye, the stereoscopic view of the 3D image data is obtained. Thus, the 3D display provides a stereoscopic display at a plurality of viewing positions.

  It will be appreciated by those skilled in the art that two or more of the above-described embodiments, implementations, and / or aspects of the invention may be combined in any manner deemed useful.

  Modifications and variations of the apparatus, 3D image data, methods, and / or computer programs corresponding to the described modifications and variations of the apparatus and / or 3D image data can be implemented by one of ordinary skill in the art based on this specification. Further preferred embodiments of the device and method according to the invention are given in the appended claims. The disclosure of the claims is incorporated herein by reference.

  These and other aspects of the invention will be apparent from and will be elucidated with reference to the embodiments described hereinafter.

Fig. 2 shows a device for processing 3D image data, a 3D display for displaying 3D image data and a viewer conflict detector. A top view shows a 3D display and two viewers. Indicates indication of viewer position conflict. Fig. 2 shows an autostereoscopic display based on a light steering backlight with a plurality of light sources. A collision section for detecting a viewer position beam conflict is shown. 3D shows a 3D display device and viewer menu. 1 shows an apparatus having a 3D display, a broadcast receiver, an Internet receiver, a storage medium and a storage medium reader. A method for detecting viewer conflicts is shown.

  The figures are just diagrams, not to scale. In the figure, elements corresponding to elements already described are given the same reference numerals.

  FIG. 1 shows an apparatus 120 that processes 3D image data 122. The device has a 3D display 160 that emits at least two views of 3D image data. The views perceive a first of the at least two views via the left eye and a second of the at least two views via the right eye to provide a 3D effect. Autostereoscopic viewing of 3D image data at a plurality of viewing positions 182 and 184 by a viewer. It is known that at unfavorable positions, the views can overlap and the eye can receive contributions from different views, called so-called crosstalk.

  The 3D image data 122 may be stereoscopic image data having a left view and a right view, or so-called multi-view image data, i.e., having image data for each or a subset of the series of views 100. Although not shown in FIG. 1, the apparatus 120 may have a view renderer for generating 3D image data 122 for this purpose, eg, from 2D image data and depth data, or from stereoscopic image data. The view renderer may be a view renderer as described in WO 1997/023097 (A3).

  The 3D display 160 may be an autostereoscopic 3D display and is shown in FIG. 1 as a combination of a display layer 162 and an optical layer 164. The display layer 162 may be, for example, a liquid crystal display (LCD) layer or an organic light emitting diode (OLED) layer. The optical layer 164, eg, a lenticular lens, provides multiple views with at least one conical field of view.

  The optical layer 164 may be a so-called lenticular lens array layer or a parallax barrier layer. Such and other examples of optical layers are known in the field of autostereoscopic 3D displays. Display layer 162 and optical layer 164 cooperate to emit adjacent views 100 of 3D image data 122 in conical field of view 180. FIG. 1 shows a single cone field 180 emanating from the center of the 3D display 160. It will be appreciated that the display layer 162 and the optical layer 164 cooperate to emit a conical field of view from multiple locations located across the display surface of the 3D display 160. For example, if the optical layer 164 has a plurality of microlenses, each of the plurality of microlenses may emit a cone field that is the same as or similar to the cone field 180. However, for illustration purposes, FIG. 1 shows only a single conical field of view.

  The conical field 180 has a series of views 100. FIG. 1 shows a top view of 3D display 160 and conical field of view 180. Thus, the series of views 100 has a rightmost view 10 to a leftmost view 90. Each view of the series of views 100 is shown as being approximately equally distributed within the conical field of view 180, ie, emitted with substantially equal solid angles relative to the origin of the conical field of view 180. It will be appreciated that other suitable distributions of the series of views 100 within the conical field 180 may also be possible. Note that the assignment of what is the leftmost view is a choice. As such, the location of the cone transition depends on the respective view rendering at the corresponding pixel.

  The viewer positioned in the space in front of the 3D display 160 observes that the viewer's left eye observes the left view of any two different views of the series 100 and the viewer's right An autostereoscopic view of the 3D image data 122 can be obtained by observing the two different views by observing the right view of the two views. This provides a plurality of viewing positions within the conical field 180 where autostereoscopic viewing is provided. Note that the 3D display 160 may be optimized to provide autostereoscopic viewing to a viewer who recognizes two adjacent views in a series of views instead of any two different views. Will be done.

  Alternatively, the 3D display may be a sequential display, also known as a light steering display, that directs a light beam to each eye of the viewer, as will be described later. For the light steering system, the back unit 162 constitutes a light steering backlight that generates a directional beam of light, and the front panel 164 is an LCD panel. In such a device, the optics are behind the LCD layer.

  Thus, for purposes of explanation, the present invention will now be described with respect to the above two adjacent views in a series of views, i.e. viewing positions where the viewer perceives each sequential view.

  FIG. 1 shows two examples of viewing positions: a central viewing position 182 near the central axis of the conical field 180 and an outermost viewing position 184 near the side or boundary of the conical field 180. Show. In both of these viewing positions, the viewer perceives two adjacent views in the series of views.

  It is known that the term “viewing position” is understood to refer to all positions where two adjacent views of a series of views can be recognized. Thus, for example, the central viewing position 182 is a viewer within the conical field 180 where the left eye recognizes the view 50 of the series of views 100 and the right eye recognizes the view 40 of the series of views 100. Is understood to refer to all possible positions.

  The apparatus comprises a processor 140 that processes 3D image data to generate the above-mentioned at least two views for display on a 3D display. The apparatus further includes a viewer detector 130 that detects the viewer position of the viewer in front of the 3D display. The viewer detector 130 may include a camera that acquires an image of the space in front of the 3D display, and may include detector software that processes the image to detect the viewer's face. Other head tracking systems may also be used and are known as such.

  The processor 140 further includes a viewer conflict detector 141 arranged to detect and resolve a viewer conflict. Initially, the unit obtains at least a first viewer position of the first viewer via the viewer detector 130. The unit then detects a viewer position conflict at the first viewer position where the first view and the second view do not have a 3D effect on the first viewer. Finally, the unit controls the generation of views by the processor 140 in response to the detected viewer position conflict. In addition, the unit changes at least one of the above-mentioned at least two views as perceived by the first viewer.

  FIG. 2 shows a 3D display and two viewers in a top view. The 3D display 200 is an array of pixels, eg, an LCD with five cooperating subpixels to provide five views in a conical field, where the subpixels and views are numbered 1-5. Yes. The conical field of view is repeated, and the second conical field of view is shown at the bottom of the first conical field of view by repeated view 5 (the number “5” font is different). Thus, if the viewer's eyes are within a single cone, the viewer will see a good 3D 3D. If one eye is outside the main cone and inside another cone, the viewer sees a bad 3D (also called reverse vision, anti-stereo).

  Thus, when a viewer moves in the horizontal direction, the viewer recognizes a different set of adjacent views. The viewer position is detected by a viewer detector (not shown) and presumed to be known in the system for the position of the conical field of view. The first viewer 210 is detected upon receiving views 3 and 4 of the first conical field of view that provides an accurate 3D effect. The second viewer 220 is detected upon receiving a first conical field view 1 and a second conical field view 5 resulting in an inaccurate 3D effect of reverse vision.

  When viewing autostereoscopic 3D TV, there is a position in front of the TV that results in good 3D (stereo) and poor 3D (anti-stereo), also called reverse vision or inaccurate 3D effects. While watching a movie, the viewer tends to move somewhat and may (partly) enter the anti-stereo area. For the viewer, it is not always obvious that the viewing position is not perfect, which can cause eye strain or worse. It is therefore proposed to provide active feedback to the viewer when the viewing position is inappropriate. However, other viewers in the correct viewing position should not get this feedback. Detect when one viewer is in or enters an anti-stereo area and provide a feedback cue in the image to the viewer that the viewer receives an accurate 3D effect It is suggested to inform you that you cannot. In the example, the position of the second viewer 220 has a conflict because the second viewer 220 has received two outer views from two different conical fields of view. An indication is proposed that works only for viewers who are at the boundary of the conical field of view but not for viewers already sitting in the center of the conical field of view.

  FIG. 3 shows an indication of a viewer position conflict. The figure has a 3D display as shown in FIG. The first viewer 310 is detected upon receiving views 4 and 5 of the first conical field of view that provides an accurate 3D effect. The second viewer 320 is detected upon receiving a first conical field view 1 and a second conical field view 5 resulting in an inaccurate 3D effect of reverse vision. Therefore, the second viewer 320 has a viewer position conflict. At this time, the system generates a warning message 330, for example, the text “bad location” or similar warning, or a graphical warning symbol. The warning message is inserted into view 1 as long as the second viewer 320 is in the conflict position. Optionally, the system may also insert a second warning message, eg, second warning text 340, into the recognized second view 5. Note that the system first detects whether any of the views 1, 5 that are recognized at the conflict location are also recognized by at least one other viewer. In the example, the first viewer 310 also recognizes the view 5. Thus, the system generates and inserts only the first warning message 330 in view 1. However, if the first viewer receives views 3 and 4 as shown in FIG. 2, both warning messages are inserted in views 1 and 5, respectively.

  Thus, when the viewer enters the anti-stereo area, the text message is superimposed on at least one of the images perceived by the viewer, for example, views 1 and 5. The warning is only visible to this viewer 320 and appeals to the viewer 320 to exit anti-stereo until the viewer 320 enters the central region of the conical field of view (view 2-3-4). When the viewer 320 enters the central region of the conical field of view, the text message is turned off and the viewer 320 is completely free between views 1-5 until the viewer 320 crosses the cone boundary again. Can move on.

  As explained above, text messages or graphical symbols can be given only to one eye. For example, if the first viewer views the correct stereo through views 4 and 5 and the second viewer views the anti-stereo through views 5 and 1, the text message is only in view 1. May be displayed, thereby not disturbing the first viewer.

  There are other options for giving the viewer feedback to exit the anti-stereo section instead of text messages superimposed on the image. In embodiments, the color of the image is changed, eg, inverted. The color image may also be replaced by a black and white image. In further embodiments, the image content may be changed. For example, as long as the viewer is at the viewer conflict position, a black screen or a red screen may be temporarily displayed. A further alternative is to show the flickering text or symbol and get the most attention. For example, the symbol may be an arrow overlay that indicates a preferred direction of movement to escape the conflict.

  In an embodiment, a viewer at the conflict location may be provided with a 2D image instead of a 3D image by generating two views with the same content. For example, in the example shown in FIG. 3, view 1 is used only by the right eye of the second viewer 320. Thus, the content of view 1 may be copied from view 5. Therefore, since both eyes of the second viewer 320 receive the same image, the 2D (mono) effect is recognized. The first viewer 310 still receives the 3D effect because views 4 and 5 are not affected.

  FIG. 4 shows an autostereoscopic display based on a light steering backlight with a plurality of light sources. In the display, the light sources 411 and 421 are arranged behind the lens 403 in the backlight unit 410, and each light source corresponds to a specific direction. In front of the lens is an LCD display 405. It should be noted that a single lens may illuminate multiple pixels in the LCD, while a directional beam is generated towards the viewer's particular left or right eye. The LCD sequentially displays a left view and a right view. Some light sources may be activated during the time that the left view is displayed to direct each beam to the viewer's left eye. The corresponding light source is activated when the display displays the right view at the next time.

  FIG. 4 represents a sequential light steering system with two simultaneous viewers 410, 420. Each viewer is provided with a stereoscopic 3D image by presenting left and right images in time sequence. Here, it is assumed that the LCD has a refresh rate of 240 Hz. In that case, the LCD displays the left eye image for 1/120 second as indicated by the solid line from the light source to the eye. During this interval, the light source corresponding to the left eye direction of the viewers 410 and 420 is turned on. In the next 1/120 second, the right eye image is displayed when the light source corresponding to the left eye direction is turned on, as indicated by the broken line from the light source to the eye. By repeating this process in time sequence, the viewer sees a black frame in the left eye and an image in the right eye and a black frame in the right eye alternating with the image in the right eye in 120 seconds.

  The display in FIG. 4 uses spatial multiplexing of the light sources in the backlight to accommodate different viewing directions. In the example, the spatial separation between viewing directions is sufficient to provide stereoscopic quality 3D for both viewers. Note that when the right eye of the first viewer 410 is at the same angle as the left eye of the second viewer 420, only one viewer can be supplied with stereo. At angles close to this situation, crosstalk may be visible by either or both viewers. As a result, it is virtually impossible to provide good stereo quality for both viewers simultaneously. Such a position of the two viewers is called a viewer position beam conflict.

  There are various ways to handle this situation. If no precautions are taken, both viewers will experience an undesirable increase in crosstalk. It is proposed to switch at least one viewer from stereoscopic to planar view to avoid bad quality. This can be achieved by presenting only the left image or only the right image of the stereo content. That is, all backlight sources corresponding to the positions of both eyes of both viewers are turned on when showing the left (or right) image, while in the next field the corresponding right (or (Left) The image is not shown by turning off the backlight source.

  FIG. 5 shows a collision interval in which a viewer position beam conflict is detected. The figure shows a sequential type display similar to FIG. With the set of light sources 411, 421, each directional light beam for a particular direction of each viewer at a particular viewing position can be generated. In the figure, each viewing section 515, 525 is shown to represent a viewing position where the corresponding directional beam can achieve the desired 3D effect. In the system corresponding to this embodiment, the viewer conflict detector is arranged to obtain the first and at least second viewer positions of the second viewer via the viewer detector. . Their location may be represented by viewing segments 515, 525. Thereafter, the viewer conflict detector is arranged to detect a viewer position beam conflict in which one view at the first viewer position coincides with one view at the second viewer position. Is done. In the figure, the viewer viewing sections at such colliding viewing positions partially overlap.

  A method of determining whether or not to switch from the stereoscopic view to the planar view is to define a viewing section 515, 525 or a collision section that is a fixed range around the viewer. FIG. 5 represents the use of a V-shaped collision section. The idea is to present stereo content to both viewers as long as the collision ranges do not overlap. When the ranges overlap, at least one or both viewers are presented in the planar view mode.

  An extension of this embodiment is to take into account viewer movement. Assume that person B sits on the couch and is not moving. On the other hand, the person A is walking left and right, and the specific movement passes through the viewing direction of the person B. In this case, it is natural to give the person B a higher priority in receiving good image quality. The reason is that a much larger variation than that of the person A appears on the screen viewed by the person B. In addition, degradation of image quality during movement is considered acceptable with high certainty. Finally, it is technically more difficult to track a moving person, which makes the fallback scenario stereoscopic anyway attractive. Similarly, other clues for care can be used. In the case of a person who is sitting and moving, it is natural to prioritize the vertical position segmentation, i.e. sitting position conflict.

  In a further embodiment, if the person A approaches the person B but suddenly decides not to leave the person B's collision section again, both viewers are switched to planar view. Here, it can be considered that the person A is seated next to the person B. In this case, it is not desirable for the viewer A to still receive an image distorted by crosstalk. This problem can be eliminated by setting a threshold on a timer that records the time that overlap exists between collision intervals. This time delay is also excellent in making the system insensitive to noise. Otherwise, it can introduce random frequent switching when the persons are close to each other. It is natural that the stereoscopic mode is enabled again when the person is away and the collision sections do not overlap. It is proposed to wait for scene transitions to minimize frustration introduced by switching. In addition to the timer, the scene transition may also be used to trigger a transition back to the planar mode or back to the stereoscopic mode.

  FIG. 6 shows a 3D display device and a viewer menu. The 3D display device 610 gives an example of a 3D television unit. A menu is shown on the display screen. The left side of the menu schematically shows the two viewers currently selected when receiving image data in 3D mode. The right side of the menu shows the viewer currently set when receiving the 2D version of the image data. The menu can allow the setting for each viewer to be changed. The graphical display of the menu makes it possible to control viewer-specific 2D / 3D settings. TV shows a close-up of all viewers. With remote control, viewers can be divided into two groups.

  Embodiments allow other types of interaction between viewers who prefer 3D. By combining the head tracking unit with face recognition, it is possible to distinguish different viewers. In this way, for example, while another viewer receives the stereoscopic content, it is possible to present a planar image to a specific viewer. FIG. 6 shows an example of an on-screen menu that can be used to group detected viewers in a planar group and a stereoscopic group. The same approach can be used, for example, to prevent a child from seeing in stereoscopic 3D mode. Observed depth for stereo-like content decreases with distance from the screen, so it is natural to switch to planar view mode for viewers that are too close to the screen. This may also be used when the viewer is too far away. In this case, the angular distance between eyes becomes so small that crosstalk cannot be avoided. Optionally, it is possible to include additional views rendered at different depths by using a distant display. By selecting paired views for different views, it is possible to provide different amounts of depth for different viewers. For example, the system may select an appropriate set of views for each individual viewer.

  As described above, the situation where a plurality of viewers are located at the same outer corner with respect to the display is examined and improved. Additional user specific 3D settings based on face recognition have been described.

  The proposed system applies to a class of autostereoscopic displays that use head or pupil tracking to provide images to the viewer's left and right eyes. More specifically, embodiments are a class of autostereoscopic displays that use backlight and light steering to direct light to each of the eyes. An exemplary configuration is to use multiple switchable light sources behind the lens array to direct the light. Examples of such displays can be found in the literature available at http://www.seereal.biz/en/autostereoscopy/papers/EI08%206803-24%20Web.pdf: Hagen Stolle, Jean-Christophe Olaya, Steffen Buschbeck, It is discussed in Hagen Sahm, Dr. Armin Schwerdtner, Cialial Technology (Germany), “Technical solutions for a full-resolution auto-stereoscopic 2D / 3D display technology”. The head tracker is used to define the viewer's one or method for the screen. By addressing the appropriate light source, light is directed to either the left or right eye. In this way, it is possible to provide a viewer with a stereoscopic 3D image in time sequence. Further examples of view sequential 3D displays can be found in the literature available at http://download.microsoft.com/download/D/2/E/D2E425F8-CF3C-4C71-A4A2-70F9D4081007/Backlightforviewsequentialautostereo.docx: Travis, N Emerton, T. Large, S. Bathiche, “The design of backlights for view-sequential 3D”.

  With spatial multiplexing of light sources, light steering can be used to provide stereo-like 3D images to multiple viewers simultaneously. This is at least possible if the viewer is located in clearly different directions relative to the screen. However, if the directions to the viewer are too similar, it is no longer possible to address the left and right eyes individually. More specifically, the light directed to the right eye of one viewer begins to partially overlap the light of the left eye of the other viewer. Or vice versa. This causes “crosstalk” and thus degradation of 3D performance. In the worst case, one of the viewers sees anti-stereo.

  Another type of interaction that has been identified for the case of multiple viewers is either 2D or 3D viewing preference. Stereopsis does not deal with all visual stimuli available in the real world. Conflicts between different depth cues can result in visual discomfort. Some people are more sensitive to this than others, so their preferences for 3D can be different. Also, the artifacts introduced by the system can give rise to different preferences. Finally, individual depth perception depends on their distance to the screen. Viewers close to the screen may perceive depth more than intended.

  Since it is impossible to achieve good 3D performance for viewers at all arbitrary locations with the same projection, the system described above detects the location of the collision and then viewers in the same direction Provide a plan view or warning signal to a part of Conflicts between different preferences when viewing content in 2D or 3D can be eliminated by combining head tracking with a face recognition scheme.

  7 shows a 3D display 160 that displays 3D image data 122, a broadcast receiving unit 710 that receives 3D image data from a broadcast transmission 712, an Internet receiving unit 714 that receives 3D image data from the Internet 716, and 3D image data. Is shown from the storage medium 720 and the storage medium reading unit 718 is shown. It will be appreciated that the apparatus 700 may have only some of those components instead of having all of the components described above. The device 700 may be, for example, a 3D television having a multi-view 3D display 160 based on lenticular. The apparatus 700 may further include, for example, a Blu-ray player, a Blu-ray recorder, a set top box, a personal computer, a hard disk recorder, and the like. Although not shown in FIG. 7, the apparatus 700 allows the user to instruct the viewer conflict detector 140 to turn on or off the detection of viewer conflicts or to set some user preference. User input means may be included. The user input means may be a remote control signal receiver that allows the user to instruct the viewer conflict detector 140 using a remote control.

  Broadcast receiver 710 may be of any suitable type for receiving terrestrial, satellite or cable broadcasts, for example. The Internet receiver 714 may again be of any suitable type, for example, having modem functionality required by ADSL, Ethernet, WLAN, UMTS, etc., or an interface protocol such as TCP / IP may be used. Storage medium reader 718 may be of any suitable type for reading 3D image data 122 from storage medium 720. The storage medium 720 is of any suitable type, such as Blu-ray, DVD, flash memory, ROM, RAM, etc. The storage medium may store a computer program that executes the method for viewer conflict detection as described above when loaded into the device.

  FIG. 8 illustrates a method for detecting viewer conflicts. The method starts at START and processes 3D image data for a 3D display device as described above. The apparatus includes a viewer detector that detects the viewer position of the viewer in front of the 3D display. The method then processes the 3D image data to generate at least two views for display on the 3D display. The method further comprises detecting a viewer conflict. Detection begins with step OBTVP, which obtains at least a first viewer position of the first viewer via a viewer detector. Next, step DETVC detects a viewer position conflict at the first viewer position where the first view and the second view do not have a 3D effect on the first viewer. If no viewer conflict is detected at decision step VC, the method continues to obtain the viewer position at step OBTVP. If a viewer conflict is detected in decision step VC, the method changes at least one of the at least two views received by the first viewer to change at least one of the views in step MODV. Continue by changing. If no further 3D image needs to be supplied, the method ends at step END. It is known that a previous change is removed if the method has detected that the view has been previously changed and the previous viewer conflict no longer exists.

  It will be appreciated that, for clarity, the above description has described various embodiments of the invention with reference to different functional units. However, it will be apparent that any suitable distribution of functionality between different functional units or processors may be used without departing from the invention. For example, functionality represented to be executed by separate processors or controllers may be executed by the same processor or controller. Thus, a reference to a specific functional unit should not be considered a strict logical or physical structure or system, but merely a reference to an appropriate means of providing the described functionality. .

  The invention can be implemented in any suitable form including hardware, software, firmware or any combination of these. The invention may optionally be implemented at least partly as computer software running one or more data processors and / or digital signal processors. The elements and components of an embodiment of the invention may be physically, functionally and logically implemented in any suitable way. In practice, the functionality may be implemented in a single unit, in multiple units, or as part of other functional units. As such, the present invention may be implemented in a single unit or may be physically and functionally distributed between different units and processors.

  Although the present invention has been described in connection with some embodiments, it is not intended to be limited to the specific form set forth herein. Rather, the scope of the present invention is limited only by the accompanying claims. In addition, while features may appear to be described in connection with particular embodiments, those skilled in the art will recognize that the various features of the described embodiments can be combined in accordance with the present invention. I will. In the claims, the word “comprising” does not exclude the presence of other elements or steps.

  In addition, if individually listed, a plurality of means, elements or method steps may be implemented by eg a single unit or processor. In addition, although individual features may be included in different claims, they may be advantageously combined in some cases and inclusion in different claims implies that a combination of features is not feasible and / or advantageous. It is not what you do. Also, the inclusion of a feature in one category of claims does not imply a limitation to this category, but rather indicates that the feature is equally applicable to other claim categories as needed. In addition, the order of features in the claims does not imply any specific order in which the features should function, and in particular, the order of the individual steps in a method claim should be performed in that order. It does not imply that. Rather, the steps may be performed in any suitable order. In addition, singular references do not exclude a plurality. Thus, reference to “one”, “first”, “second”, etc. does not exclude a plurality. Reference signs in the claims are provided merely as a clarifying example and shall in no way be construed as limiting the scope of the claims.

Claims (8)

Emits more than two views of the three-dimensional image data, the views through the left eye through the first view of the plurality of views and through the right eye to produce a three-dimensional effect. to allow autostereoscopic of the 3-dimensional image data in a plurality of viewing positions by the viewer to perceive a second view of the plurality of views Te, and 3D display,
A processor that processes the 3D image data to generate the plurality of views for display on the 3D display;
A viewer detector for detecting the viewer position of the viewer in front of the three-dimensional display;
The processor is
Obtaining at least a first viewer position of the first viewer via the viewer detector;
Detecting a viewer position conflict at the first viewer position where the first view and the second view do not provide the three-dimensional effect to the first viewer;
Obtaining at least a second viewer position of the second viewer via the viewer detector;
Select one of the plurality of views received by the first viewer to indicate to the first viewer that the first viewer is not subjected to the three-dimensional effect. To generate the view depending on the detected viewer position conflict, so that the one selected view is not a view recognized by the second viewer. A display device further comprising a viewer conflict detector arranged to control. The three-dimensional display is a multi-view display arranged to emit adjacent to a series of more than two views of the three-dimensional image data in a conical field of view, the series of views comprising a plurality of views in the conical field Enabling autostereoscopic viewing of the three-dimensional image data at the inning position, and the viewer conflict detector,
Viewer position at the first viewer position, wherein the first view is a first cone field from a first cone and the second view is a second cone field from a second cone. Arranged to detect cone conflicts,
The display device according to claim 1. When the viewer conflict detector detects the viewer position cone conflict,
Overlaying graphical elements in each conical field of view,
Overlaying a text message on each conical field of view,
Changing one of the first cone field and the second cone field by at least one of changing the image content of each cone field by temporarily showing a black screen or a red screen Arranged to control the processor,
The display device according to claim 2. When the viewer conflict detector detects the viewer position cone conflict,
Changing the brightness of each conical field of view,
Arranged to control the processor to change one of the first cone field and the second cone field by at least one of changing the color of each cone field;
The display device according to claim 2. The three-dimensional display is arranged to emit a left view and a right view in succession via each first directional beam to the viewer's left eye and each second directional beam to the right eye. A sequential view display, wherein the view enables autostereoscopic viewing of the three-dimensional image data at each viewing position by combining respective directional beams, and the viewer conflict detector comprises:
2. The viewer position beam conflict of claim 1 arranged to detect a viewer position beam conflict in which one view at the first viewer position coincides with one view at the second viewer position. Display device. The viewer detector is:
Face detection in video images from cameras,
Distance information from the distance sensor,
User input from a remote control unit operated by the viewer at the viewer position,
Arranged to detect the viewer position of the viewer in front of the three-dimensional display based on at least one of detecting a viewer tagged object positioned by the viewer at the viewer position To be
The display device according to claim 1. A method for processing 3D image data for a display device, comprising:
The display device includes:
Emits more than two views of the three-dimensional image data, the views through the left eye through the first view of the plurality of views and through the right eye to produce a three-dimensional effect. to allow autostereoscopic of the 3-dimensional image data in a plurality of viewing positions by the viewer to perceive a second view of the plurality of views Te, and 3D display,
A viewer detector for detecting the viewer position of the viewer in front of the three-dimensional display;
The method includes processing the 3D image data to generate the plurality of views for display on the 3D display;
The method is
Obtaining at least a first viewer position of the first viewer via the viewer detector;
Detecting a viewer position conflict at the first viewer position where the first view and the second view do not provide the three-dimensional effect to the first viewer;
Obtaining at least a second viewer position of the second viewer via the viewer detector;
Select one of the plurality of views received by the first viewer to indicate to the first viewer that the first viewer is not subjected to the three-dimensional effect. To generate the view depending on the detected viewer position conflict, so that the one selected view is not a view recognized by the second viewer. The method further comprising detecting a viewer conflict by controlling. A computer program comprising instructions for causing a processor system to perform the method of claim 7.

Images